Medical imaging is one area that is core in the diagnosis and treatment of various medical conditions. However, con-ventional 2D medical imaging, including Magnetic Resonance Imaging scans, does not provide the level of depth per-ception required to effectively interpret 3D anatomy from 2D views. This can be seen in conventional MRI scans that are widely used. This article presents a new method that enables the inference of depth information directly from a 2D MRI image using a monocular depth estimation algorithm called MiDaS, which is a state of-the-art monocular depth estimation model that can perform this operation. This method starts with some general image processing steps, which include uniformly scaling images to their correct dimensions and uniformly setting the brightness levels of their pixels to ensure consistency and modifying image colours to a format that works well with image processing demands. This work showcases how AI-driven depth estimation can make a significant difference in traditional imaging methods by offering interactive 3D visualization. Deep learning-based depth inference combined with conventional medical imaging could hold great promise for extending radiological practice, enabling health professionals to derive deeper insights into ana-tomical structures without additional imaging modalities or increased radiation exposure.

Medical Imaging, MRI, 3D Volumetric Reconstruction, MiDaS, Image Processing, Otsu Thresholding, Marching Cubes, Depth Estimation